Question: has anyone applied Arrow's theorem to derive design constraints on distributed AIs? Kinda like a CAP theorem for AIs.
Conversation
Replying to
seriously? do you have refs? asking for a work related thing, so actually want to know...
1
1
Replying to
yes here is my peer-reviewed journal article
Quote Tweet
does anyone really think the class "humans" can have a coherent anything, much less coherent extrapolated volition
1
2
Replying to
if by "helping" you mean "laughing maniacally and taking another bong hit" then yes I am your banana
1
4
I would write about this but somebody told me to do a self-review for july
1
1
also whenever I write about AI friendliness it tends to come out pretty retarded because I know how big & deep the field is
1

